What Can’t Moocs Teach?

Also irreplaceable: laboratory based courses, any course that requires materials or equipment not commonly found at the student’s home. Or is this too obvious to mention?

Generally valid points, but consider:

1) The median science and engineering student does not take more that one year’s worth of humanities and social sciences classes. If we grant Jennifer Morton her entire argument, it still leaves about three years worth of classes that could be potentially more accessible to disadvantaged students via MOOCs.

2) I fully agree that a form of offline social interaction is crucial for learning. I don’t disagree much with Randall Collins when he sees “ritual chains” virtually everywhere — and learning does have its rituals that require local social interaction.

However, God did not make it a Law of Nature that the only possible form of learning via social interaction is the classroom. In fact, the professors themselves don’t go to classes in order to perform the social ritual part of learning — professors go to conferences (I suppose professors can also learn from students, but that only goes so far.)

There are other empirical examples of social interactions outside the classroom that have excellent results on the learning metric: the various national and international Olympiads. It’s true that the students participating in the Olympiads are mostly self-selected and are among the best. Nevertheless, there is a huge amount of learning done by participating students and they learn material that goes much further than anything taught in the classroom.

Given those successful examples, I’m moderately optimistic that regular online interaction (e.g. discussion threads) coupled with periodic, shorter, but intense offline social interactions, can be a successful model for teaching science and engineering MOOCs.

I think this experiment is very interesting.

http://harvardmagazine.com/2012/03/twilight-of-the-lecture

http://www.npr.org/2012/01/01/144550920/physicists-seek-to-lose-the-lecture-as-teaching-tool

This makes you question the primary structure of a MOOC, where a lot of learning takes place by watching passively.

I also think this shows that experimentation is critical in teaching. Many of the people I know are engaged in a long term process of improving teaching via experimentation and innovation. What kind of questions work? What kind of discussions work? And so on. Understanding the culture of one’s own college and university is also critical to teaching.

As someone very familiar with online discussion and who enjoys it, I question whether developing critical thinking and speaking skills will as readily be the result of online discussion as of classroom interaction. There are better and worse ideas (and sometimes only one good idea) that a professor is trying to move students toward and online discussion often moves toward consensus, or will get derailed by misinformation. The professor’s role as guide will be very watered down and you can’t have a fully engaged, fully collective discussion because it tends to take place in terms of replies, which are one on one.

There are so many ways MOOCs can go awry but what I find most troubling is that MOOCs are proposed for the hoi polloi–not the elite students, who are still deemed worthy of contact with real experts.

“Are you implying that engineering classes in particular don’t require intense personal interaction and practice therein? I would respectfully disagree with that analysis.”

I do think that face to face interaction is less important in STEM fields, yes.

That is something different than collaboration not being important (you can collaborate online), and I think periodic face to face interactions are helpful. But they are not needed to the degree that they are needed in humanities.

Cranky,

A more expansive answer would be helpful. What are the specific needs of engineering learning (what kind of engineering, exactly?) that require continuous face to face interaction that can not be replaced by online interaction (including video chat)?

The current economic system doesn’t produce enough opportunities to justify sending so many people to college. Since simply cutting back on enrollments would require admitting what’s really going on, MOOCs provide a convenient way to pretend that every one can get a college education that leads somewhere. A pasifier, however, is not quite the same thing as a breast or even a bottle, even if it does shut the kid up for a while.

I’m a bit puzzled the idea of college as a place to pick up middle class social norms. I must’ve gone to the wrong college (it was a largish fairly decent public school). The years I spent there involved a very rapid erosion of the middle class social norms I had acquired from my middle class parents. It took many years away from that environment to recover a semblance of the attitude expected amongst the bourgeoisie.

Having done this myself (in order to have some background for a conference presentation on MOOCs, I would encourage anyone who is interested in the promises and pitfalls of MOOCs for higher education *to take one* and see how it actually works. What I took away from the one I took was (a) the course content, as prepared and presented by the instructor, was equivalent to what I would have expected to find in an on-campus course; (b) the discussion forums were more civilized, but less intellectually interesting, that would be optimal; (c) participation (especially in the discussion forums) was extraordinarily low; and (d) the assessment of learning (peer evaluation of written assignments) was wholly inadequate (too little of the course material could be covered in the assessments–there were only 3–and too many students had too little background in performance evaluation, and too little guidance within the course, to do it well).

But I think MOOCs will survive, and evolve, in ways that are likely to make them at least somewhat effective, if only as a substitute for large lecture sections (takling heads with power-points in a classroom aren’t all that hard to replace with talking heads with power-points on a computer screen). Here’s what I think it will take: A strong on-campus component. By which I mean (1) a faculty member on campus who supervises the course, designs the course learning objectives, prepares the assignments, and supervises course assistants; (b) discussion sections or on-line discussion forums with course assistants (read teaching assistants) as active participants; and (c) on-campus grading of all the assignments. The latter two of these three conditions will require adequate training of the course assistants wo be effective.

And, doing it this way won’t save a boat load of money.

MOOCs may also have some role as supplemnetary material in other classes.

And see this for a discussion of what I am now calling SPOCs–small, private on-line classes: http://chronicle.com/blogs/wiredcampus/3-universities-will-grant-credit-for-2us-online-courses/45143

We’re being asked to accept the idea that the university — as opposed to the internet — provides the opportunity to argue? HMM.

Anyway, certainly, you can’t learn eye contact conventions on the internet. But it goes both ways. You can’t learn how to collaborate via mailing lists and bugtrackers in a meatspace discussion group.

Donald, I concur with what you took away from the MOOC experience. I enrolled in the same course from down under. Whilst the lecture presentations were first rate, the additional context for deep learning were thoroughly lacking (discussion forums, assessment process). I have a friend who convenes a large open university course (distance + online learning) and his experience is that he has become more of an administrator of a platform than a teacher/researcher.

my guess is that MOOCs will unbundle the relationship between teaching undergrads and research work. That will be more of a change to colleges than anything else.

My guess is that universities lower down the food chain will seize on MOOCs as a way to do just this, drastically reducing the number of academics employed in the process.

My reaction to MOOCs is frankly Luddite. It seems to me that they threaten my job and my friends’ jobs, and I don’t like ’em.

I agree that many college classes in the status quo aren’t taught that well. But I don’t see how the solution to large lecture classes is to put the lectures online. Lecture is the least effective way to teach, and well-taught classes rely less on lecture and more on problem sets, group work, reading, discussion, papers, and lots of feedback, and most of these can’t be reproduced online (I have done online courses and the “discussions,” such as they were, were not as rigorous as even a comments section of a blog).

So I disagree that it’s only small discussion classes that can’t be reproduced online. Feedback from professors is key, and the way these online classes are run there is very little of it for students. In the best class I took this past year had no discussion, but the professor had us writing 5 large proofs a week and he was reading them all, commenting on every step that didn’t make sense, and I cherish that feedback. There is no online replacement for that (unless a professor marks up a Word document, but that’s less efficient than using a pen and paper).

Also, supervision is key in traditional classes. 18- to 22-year-olds are still kids. They need weekly homework and to be told to turn in their assignments and to have fellow students to keep them going. They need regular tests and quizzes so they don’t leave everything to the last minute. This is my theory about why online schools have such terrible graduation rates – no supervision.

On the other end of the spectrum is lecture. Lecture is about as stimulating as watching TV (as MIT found last year) and any actual learning from straight lecture will be from notes being reread later. So why do MOOCs focus so much on lecture when lecture sucks anyway? Is it just because it’s the first thing people who don’t understand education think of when they think about education? And it’s cheap to put online? There is some good software for learning basic math and econ and other basic subjects out there that does teach (infinite practice problems because the computer can just switch out the numbers), but that’s been available for years while the MOOC obsession is more recent.

With that in mind, it’s a cute idea that administrators would use the money freed up by MOOCs to offer smaller classes where needed. Considering how tuition costs have risen much faster than professor salaries have, while professors are being replaced by barely-paid adjuncts faster than ever… yeah, I’m sticking with “cute.”

The whole MOOC thing, as it’s being done now, seems like a scam to me, to get students to fork over money for an inferior product with zero marginal cost. That they’re focusing on delivering part of the product – lecture – and the part that only people who don’t know understand how education works will think is important, just seems like a con to me. The business model depends on people not understanding what made the product valuable in the past.

Videos can’t tell you where your proof doesn’t make sense, but they can replicate what your parents think college should look like, especially if your parents didn’t go to college.

The key difference between online teaching/distance learning and MOOCs is the “Open” bit, i.e. having them for free. The Open University, University of London etc have been doing these kind of courses for years (my husband’s been involved with distance learning legal courses, for example), but it takes a lot of time and effort both to produce courses tailored for distance learning and then to provide student support remotely. And neither of those come cheap. You can get some economies of scale with designing very standardised courses, but it’s very hard to scale up student support and still keep it good quality.

I’ve discussed possible courses that MOOCs might work for in more detail elsewhere. I think they might work effectively for taster courses and as introductions for some specialist fields (e.g. rare classical languages), but not in many other contexts.

@20: That’s precisely the point. MOOCs are no different, fundamentally, from correspondence courses or learning by watching a video or reading a book. These methods haven’t been viable alternatives to a college education for most people, although they do work for some. And the basic reason is that they only serve certain styles of learning and that they are not as good for learning as interaction with an actual teacher.

This is in reply to points raised by several people.

I find it interesting that much of the support for MOOC’s is allegedly based on the idea that large lectures are bad, since MOOC’s are large lecture grown like the Blob.

In terms of ‘big data’, we do have some on MOOC’s – what’s known publicly is that the completion and pass (percentage) rates are in the single digits, and that university if California which tried them for actual courses got similar results.

We do need more data to analyze the potential of MOOCs — but we need quality data, which can come only from further experimentation. It’s highly unlikely that MOOCs will stay in the form that they are now.

It’s a bit silly to compare completion rates at this stage, when the payoff for completing a MOOC is minimal, as there are no generally accepted accreditation mechanisms for them. Even beyond the lack of accreditation, the “let’s try it out a bit, since it costs me nothing” effect should be large anyway.

Here is a classic warning about simplistic extrapolation from current data in highly dynamic environments — even when the extrapolation is done by experts.

In 1995, in an article called “The Internet? Bah!” Clifford Stoll writes:

“After two decades online, I’m perplexed. It’s not that I haven’t had a gas of a good time on the Internet. I’ve met great people and even caught a hacker or two. But today, I’m uneasy about this most trendy and oversold community. Visionaries see a future of telecommuting workers, interactive libraries and multimedia classrooms.
[…]

Baloney. Do our computer pundits lack all common sense? The truth is no online database will replace your daily newspaper
[…]

How about electronic publishing? Try reading a book on disc. At best, it’s an unpleasant chore: the myopic glow of a clunky computer replaces the friendly pages of a book. And you can’t tote that laptop to the beach. Yet Nicholas Negroponte, director of the MIT Media Lab, predicts that we’ll soon buy books and newspapers straight over the Intenet. Uh, sure.
[…]

Then there’s cyberbusiness. We’re promised instant catalog shopping—just point and click for great deals. We’ll order airline tickets over the network, make restaurant reservations and negotiate sales contracts. Stores will become obsolete. So how come my local mall does more business in an afternoon than the entire Internet handles in a month? Even if there were a trustworthy way to send money over the Internet—which there isn’t—the network is missing a most essential ingredient of capitalism: salespeople.
“

In few years everyone will be buying their pet food online. The disruptive technology of the internet makes in inevitable. Only a luddite could disagree.

Better pick up some shares of Pets.com now before you miss out!

Borders went out of business for reasons that have little to do with the Internet.

Perhaps orthogonal to this point — and perhaps moot in the youtube era — is the idea of making lecture material available at the student’s convenience. Back in the Jurassic Era when I was a student, every section of Econ 101 was taught by a full professor — via the magic of videotape. He would tape his first meeting of the day in a small classroom, not a cavernous 1000 seat lecture hall, and that same “performance” was shown to the later class meetings that day.

Seems pedestrian now but it prefigured a lot of the touted advantages of online learning: full professors teaching 100-level courses, standardized delivery and content, access to the material after the fact. This is best suited for lectures/presentation of material, not interactive examination of it, as noted above. We think of higher ed as a component of social mobility, but generally only in terms of the hallowed diploma, not for the workaday social skills that one learns. I have assumed that one of the key attributes conferred by a four-year degree is the ability to make a plan/set a goal and follow through. But I didn’t face any challenges fitting into a middle class milieu.

https://www.google.com/search?q=david+denslow&ie=UTF-8&oe=UTF-8&hl=en&client=safari

The value of a MOOC to a conventional university has little to do with the MOOC itself.

1. It showcases the syllabus of the MOOC’s offline, non-massive counterpart, demonstrating that said course has substance to it.

2. It demonstrates the value of taking offline, small scale classes precisely because of the outcome discrepancy between the MOOC and conventional course.

3. It helps squeeze scams like the University of Phoenix out of business. (Why pay umpteen K to the U of Phoenix for something MIT offers gratis? For a certificate you can show to an HR manager, who will then immediately ask the question I just did?)

4. It helps the mission of the university by spreading knowledge, however inadequately, among MOOC takers (like me) who cannot at the moment enroll in anything else.

Alex K. 07.31.13 at 2:13 pm

” We do need more data to analyze the potential of MOOCs — but we need quality data, which can come only from further experimentation. It’s highly unlikely that MOOCs will stay in the form that they are now.”

There has been experimentation. The result still s*cks. Now, I expect MOOC’s to evolve, but right now they need tremendous evolution to get to the point where they’re worth anything.

” It’s a bit silly to compare completion rates at this stage, when the payoff for completing a MOOC is minimal, as there are no generally accepted accreditation mechanisms for them. Even beyond the lack of accreditation, the “let’s try it out a bit, since it costs me nothing” effect should be large anyway.”

Please re-read my comment with comprehension.

I’ve might have mentioned here before that I’m currently running part of an online course in addition to my regular teaching.

Newsflash: doing online education right requires work, and not just during the initial setup. The folks who are under the impression that you just record the lectures and then sit back and rake in the $$$ are either delusional or grifters (or delusional grifters).

MOOC’s are proposed as a means of providing quality education at competitive, market-driven rates. Actually, they are a cover-up for decades of underfunding public universities and colleges. Sure, some faculty at elite schools are participating in the creation of MOOC’s, but I’m certain their children will continue to obtain elite, non-MOOC educations. Even if, and I dispute it strongly, MOOC’s could offer the same quality as traditional modes, students are still being tracked into an education which will exclude them from mobility into the elite. Do you think that a letter of recommendation from a MOOC instructor will carry any weight with graduate or professional schools? MOOC students will have no opportunity to socialize with faculty or graduate students where face-to-face networking provides opportunities for internships, collaboration in research or feeling comfortable in the physical presence of smart, educated people.

Alex K.: Your comments about the lesser need for face-to-face interaction in the sciences is bizarre. I can’t speak for engineering but in the biological sciences students meet with faculty during office hours to discuss ideas, ask questions and seek guidance. Same with graduate teaching assistants. Students spend a lot of time in labs. Almost all topics require hands on learning. Plant taxonomy requires handling plants, anatomy requires dissections, ecology requires field trips, physiology requires … And in all of these classes as much learning goes on among students as ever takes place in a formal lecture. Hell, when I took matrix algebra and multi-variate stat, we spent far more time working out problems together than we ever spent in class. At the University of Washington, where my wife teaches and I once did, freshman students can opt for FIG (freshman interest group) participation. Students in these groups take classes together and work together outside of class time. The results are impressive in terms of student achievement and satisfaction. The idea that the social sciences and humanities require more personal interaction than the sciences is ignorant prejudice.

The point that higher education is important in introducing students to middle and upper class mannerisms is valid. If you come from a background where everyone in your family has a college degree it may be hard to understand that things you take for granted are foreign and intimidating to students from poor or working class backgrounds. Simply having the assurance to speak up and voice an opinion in class or to a professor is a class attribute.

MOOC’s can have value, particularly for students who already have degrees and want to supplement their education without having to enroll in a new degree program. But the push for them is a cynical attempt to distract from the continued underfunding of public education.

AcademicLurker @ 43 – agreed. I have taken a few online classes, from actual, accredited non-profit universities. Some of the classes were great, other s—-d, and the difference was how much the professor was willing to put into it.

Fu Ko 07.31.13 at 5:58 am
Just to make that more explicit — the very same technological trends that are bringing in the MOOCs and making the in-person university obsolete, are also making the in-person production form obsolete by replacing it with internet-mediated production. So, the obsolete university is where people learn to collaborate in the obsolete ways…

I think that overstates the case a bit – in-person skills are still very important in many aspects of life and work. However, distance-mediated workplaces are becomeing quite common. I have been working the last couple of years with a manager I have never met, located on the other-side of the country from most of the work-group I belong to and communicating entirely by email and the occasional phone or conference call. Online classes were actually very comperable.

Online discussion groups are a bit of a crap-shoot. They really depend on the willingness and capability of class members to actually engage. I think the fact that the internet is so prone to flame-wars and the fact that you are not anonymous tends to make online-class discussions a little overly cautious. There is less tendancy to disagree.

What is nice about an online discussion is the ability to marshall facts quickly and easily, which is an important skill. If you want to disagree with what someone says about interest rate spreads and you can pull data from FRED and present a graph that makes your point, or go grab your source papers and actually back up what you say with references. This is an important skill and would be worth including in any online class, MOOC or otherwise.

Dr. Hilarius #47:

“But the push for them is a cynical attempt to distract from the continued underfunding of public education.”

From where I sit, you have the sign bit wrong on that comparison: continued perseveration on the idea that any day now someone is going to give us billions of dollars, with no strings attached, is a cynical attempt to distract from the continued need to figure out ways of educating more students per hour of faculty input.

State funding for education in the US has been declining steadily since 1975 — we’ve had 40 years to adapt, and we haven’t done it. The cost of a Bachelors’ rises every year (over a thousand percent since the late 70s) while its value is declining. Waiting around to be saved by free money is really just a way of preserving the system as it works today for tenured faculty, than try to think of a way to produce an acceptable education cheaply.

@46: Christ almighty, this reeks of self-serving bullshit. It’s those deluded professors who are frantically attempting to preserve “privilege” in your screed. As opposed to pointing out, for example, that the model being offered is an inferior substitute being promoted in bad faith?

Your long-winded post is an extended exercise in bad faith argument. People are saying that there is no fundamental substitute for teacher-student interaction. They’re saying that online education has developed a dismal actual track record for educating students; that MOOCs have dismal completion rates, serious academic integrity problems, and don’t come even close to replicating a college major program. There is substantive research that MOOCs are the worst for the least prepared students. These are the actual arguments that you’re facing, which are not ones about the format of the delivery.

And people have been talking for a very long time about the limitations and defects of the large lecture format – so complaining about attacks on it as hypocritical are absolute nonsense.

Other countries manage to provide university educations at a fraction of the cost to students that we do, just as with medical care. There is nothing fundamental that demands that we replace face-to-face education.

Ragweed, I don’t mean to suggest that meatspace is already obsolete. However, it’s at least as obsolete in the productive economy as it is in the university.

In other words, I think we should expect the proportion of education occurring through the internet instead of meatspace to roughly reflect the proportion of productive collaboration occurring through the internet instead of meatspace. The two trends are actually one trend: things are moving onto the internet.

You need tweezers for starters.

“Using only electrons, how does one build a soccer playing robot?”

As it happens, I did build a robot, for my son (or at least that was the excuse).

The way I did it was shockingly simple: I ordered the parts online and then I watched the Youtube video to properly assemble it. It doesn’t play soccer, as it has wheels, but the lesson is easy to generalize: if the materials are cheap enough people can buy them and then they can collaborate online to build whatever it is they want to build. When expensive equipment is involved then obviously MOOCs would need arrangements with brick and mortar institutions.

Surely this is not too hard to imagine?

“I haven’t seen anyone propose also banning lectures on their own campus”

Some colleges have no lectures. Are they the best positioned to navigate the higher-ed disruption?

MOOCS sound attractive, but why burden people with having to deal with video. Why not just provide a transcript? You can read a 60 minute lecture in maybe 15 or 20 minutes, so why waste all that extra time? Besides, its easier to back up and re-read than dink with rewind and the like when watching a movie. Video is inherently a very slow, inconvenient way to learn unless there is something critical that needs to be visually displayed and cannot be displayed using a more modern interactive technology.

As best I can tell, MOOCS are just correspondence courses of the kind you used to find offered on matchbook covers back when everyone smoked and matchbooks were everywhere. Judging from the 6/33 article in Fortune on ICS, the biggest provider in the last century, MOOCS have a similar completion rate to old fashioned correspondence courses. As noted in the 1933 article, the fact that so few people complete the course helps pump up the profit margin, and my guess is that this is what the big MOOC backers are counting on.

I have always learned a great deal from lectures. I miss the old planetarium lectures that have been replaced by movies and light shows. Unlike videos, it is much harder to just drift off while listening to a lecture as one can when watching videos. Besides, if you really want to understand the material, you want the traditional combination of lectures and recitations. In the science and engineering fields, the graduate students teaching the recitations can provide interesting insights, tips, techniques and hints that are otherwise impossible to discover.

It’s also interesting that the MOOC advocates don’t actually demonstrate that their techniques would give any substantive savings. (Why should they? After all, they’re like Gutenberg, obsoleting reactionaries with quills!)

Typical student to teacher ratios in universities are of order 25-1. Allot 100 K / year for salary and benefits (this is well over the median); the labor cost is about 4,000 / student/ year. Add another 1,500 a year per student for graders (advanced undergrads) and a TA.

In the best case you minimize this cost – which is a fraction of what a university education is. The real costs are the physical plant, administration, and so on. And the cheapest current forms of instruction are – you guessed it – larger lectures. (50 students, by the way, is a group size where it’s perfectly possible to have real discussions – another straw man. The actual place where MOOCs are relevant are the 100+ student classes where there is less opportunity for in class discussions – and where the savings, if any, will be small.)

This means that you actually could afford smaller class sections – if you made other trade-offs, such as less administration and less spending on the physical plant. The fact that faculty salaries have barely changed relative to inflation pretty decisively implicates these other factors as the actual things changing the cost of universities. And, of course, the defunding of public universities overall.

There is a broader context here. The MOOC advocates aren’t addressing the real costs, they’re promoting a method with serious demonstrated problems, and they’re promoting a method (effectively, barely guided self-study) that works only for a minority of students. Those are real reasons to object. (By contrast, if the alternative is “learn about science by watching TV” they can be a huge advance. The goals matter!)

“The real costs are the physical plant, administration, and so on.”

Which is why MOOCs, by unbundling actual teaching from the entire university administrative apparatus, has promising prospects for decreasing costs.

I actually agree that providing good TAs for a relatively small group of students is not that expensive — that’s why the scope for MOOCs experimenting with various online student to TA ratios, various frequencies of offline meet-ups, various certification procedures, etc. is enormous.

” The MOOC advocates [are] promoting a method with serious demonstrated problems, and they’re promoting a method (effectively, barely guided self-study) that works only for a minority of students. ”

Here’s where the completely irrational bias that treats the current form of MOOCs as the only possible form of MOOCs rears its head.

Ninety percent of start-ups fail — there is no reason why the current incarnations of MOOCs should be exceptions. Last time I took a Udacity course, which was about a year ago, they still had bugs in their grading software. Yet some people expect us to believe that all possible experimentation with using MOOCs for teaching has been exhausted. That argument is so bad that it barely deserves an incredulous stare as refutation.

I’m not at all above accusing current members of academia of selfish reasons for opposing MOOCs. I’m just fishing for more interesting good reasons for opposing them.
The fact that that established universities spend only a fraction of tuition on actual teaching is not one of those good reasons however.

@66: Good points. Online tools can be great ingredients for learning; for example, you can run simulations rather than staring at fixed static examples in a book. You can have hyperlinks to more detailed explanations elsewhere. This is a huge advance from what you get from introductory textbooks.

Most people, however, hit a place where things just don’t click, and they benefit tremendously from working closely with a teacher who can identify the stumbling block. And that is even before you go into things like writing, research, language or experiments, where the advantages of working directly with a teacher are even greater.

This seems like an effort to marry MOOC and classroom. I have no idea what to make of it yet, but the presence of Lawrence Summers and Bob Kerrey makes me queasy.

One of the things that strike me here is that a lot of the problems with learning are with overcoming preconceptions; putting people into an unfamiliar/different/”special” environment and giving them the exact same information will help here. Theatrics and formalism and “bullshit” presentation issues matter, here.

It’s weird to me that people in the sciences are saying things like “In-person interaction doesn’t matter!” I don’t know, maybe that’s so for some narrow subset of “the sciences” with which I have no personal experience, but having majored in math and physics I can tell you for a fact that personal interaction matters a great deal for learning. I can’t tell you how many classes I had where in-person interaction with the professor or a TA made a serious difference in my level of understanding. People also underestimate the effect of in-person interaction between students. At Berkeley, there was (and still is) a large room in LeConte called the Reading Room, where physics students would congregate to do their homework. Collaboration and discussion with other students was a huge part of my education; if I hadn’t had that opportunity to share thoughts with other people working on the same material, my understanding of it would have been much poorer.

In short, setting aside the other serious questions raised about MOOCs, it just isn’t the case that simply replacing in-person lectures with video lectures suffices to provide the same level of education; not in science any more than in literature or history. There’s so much more to learning than just going to lectures, and MOOCs are not going to be able to replicate those other aspects.

Jerry, “personal interaction” might be important, but the idea that personal interaction requires geographical adjacency is quite mistaken. Freenode IRC is a vastly more valuable resource (for personal interaction) than a typical university, if you are a programmer. This is not because of something to do with programming, but just the fact that programmers are the early adopters of all internet communication technologies, so there is already a critical mass of users to interact with. (Of course it’s not as if it’s just programmers, although it’s mostly people who are connected to that world in some way.)

Even at the time I was studying math at a university, and living on campus, I found it far easier to find people with whom to talk about math on IRC than on campus. I could also talk with graduate students from better universities than my own. My own university was rather poor (I had a very bad high school record) and apparently most of the math majors were looking for a career teaching primary education. They tended to have difficulty even keeping up with the classes. Only the internet provided access to the type of math student who was vying for an academic career in research (which would correspond with the type who would talk about math socially).

Perhaps at another university, my experience would have been different. But I’m not so sure. After all, the people I did find on IRC were going to (or even teaching in) those better universities themselves. If I’m talking to a math professor on IRC at 2am, it’s almost certain that he’s not available to his own students at that particular time (except through IRC!).

This in itself does not say anything for MOOCs, but the idea that there is something magical about verbal interaction that puts it above written interaction is, I think, quite mistaken. Especially computer-mediated written interaction. Having a visible log, being able to type simultaneously without negotiating interruptions, being able to search, never having to ask anyone to repeat something (basically having the computer as an extension of your memory in the immediate conversation) — these are big advantages. Even if they weren’t, the supposed advantages of being in the same room over using video-enabled VOIP are even more dubious. And yet the advantage would have to be huge, to compensate for the immense geographic restriction in the number of persons you can connect with that way.

(All that said, I will add that most of my real education, in math or anything else, has been the product of simply reading books — not personal interaction or lectures.)

“their students are learning in any case” => their students aren’t learning in any case

Didnt some nations, e.g. Sweden are already sort of get there the old fashioned way? Meaning students only spend very limited time in a classroom with few student, where the focus is on debate. Most learning then is done from books. Distance Universities are no new invention in the internet age either.

@Fu Ko, #77,

Your points about IRC are well taken. Let me just say that by no means do I want to exclude any possible avenue of learning. If someone is more comfortable reading text than they are interacting with people verbally, they should have the option of doing that; in a traditional classroom, that’s kind of the default anyway. My concern about MOOCs is that they, by design, preclude a certain kind of in-person interaction between students and teachers as well as among students.

I do strongly disagree with the notion that there’s nothing different about in-person vs. online interaction. There’s a degree of spontaneity that you get from putting lots of people together in a room that’s much harder to replicate via a medium like IRC (the round-trip time for typing and reading is much longer than for verbal communication) and virtually impossible to replicate over email. Like any scientist, I communicate via email all the time, but nothing beats showing up to a conference and actually talking to a bunch of people in your field. I had this experience just a few weeks ago when I attended a small conference and the difference between what I got out of that and what I get out of email is just staggering to me. This spontaneity is important because it allows you to learn things you might not learn otherwise, and hear perspectives which might never get brought up in an email chain because the right people aren’t on it.

Now, I’m willing to believe that this is a matter of individual difference; maybe I’m just the kind of person who learns better by communicating vocally in person than by just reading text. But I don’t think I’m at all unique in this manner. A real problem with MOOCs is that they foreclose this option for those students for whom it would be ideal; on the other hand, traditional college courses do not foreclose on the option of doing all the work by yourself in isolation (or over IRC), if that’s what works for you. In that sense, the variety of learning environments available to a student of a traditional college is a strict superset of what’s available to a MOOC student.

Clay Shirky@51: who said anything about free money? Yes, support for public education has been declining for decades. It’s not an accident or some act of nature, it is due to deliberate policy choices at the state and federal level. You seem to assume that this can’t be changed. Why? As for your snarky comment about tenured faculty, my spouse is not tenure track and my main concern is with the increased social stratification of students due to the high cost of education.

There’s no shortage of money. Just the will to allocate it to education.

I think the problem here is that people are thinking of these interactive lectures as the MOOC. To me, those are the first phase – they are what you can afford to give away free. But the economics of this are that eliminating all the expense of University but the teaching make possible to put a lot more resources into teaching.

I don’t recall taking any lecture classes at University that actually were just lectures. You had sections, which were much like seminars, a bit larger perhaps, and not expecting as much participation, but one could run them like seminars. These were typically run by grad students. Since grad students routinely did this, they were not necessarily good teachers – there was not necessarily any selectivity by teaching ability or inclination.

The interactive lectures and other materials, then, are the not the course: they are the materials. They are just (perhaps) proprietary materials of an institution.

Maybe this will become clearer if I plug some numbers in. Tuition and mandatory fees at Ivies now run over $40,000 a year. Students take 8-10 courses a year, so roughly $4500 – $5000 a course. Say I charge a fifth of that, $1000. I have the interactive lectures and other materials, but I also have small online discussion groups and individual online instruction. Out of that thousand, I give a royalty of 10% to the team that developed the materials – a star professor, maybe a curriculum/materials specialist, and a web developer (more realistically I will reduce it to a highly-flexible template). May not sound like much, but for 100,000 students that’s $10 million. And I could keep reusing this and pay royalties on it for years.

Say I break the class into sections of 15 students. Each of these get 6 hours a week of online video chat seminar, all the interactive lecture and related materials, an ongoing forum like this one (which does have advantages over real time. People can think out what they say more, and can go back and respond to precisely what someone else said, rather than how they might misremember it. Do people think these discussions would be better if we were all together in a room? I don’t. More fun, perhaps, or more bitter, but not better), and a half hour a week one on one with the section leader (if you don’t need this to master the material, this is your chance to go beyond the class material). They are graded partly on participation. The section leader gets 80% of the proceeds – $12,000. I’m under the impression that adjuncts typically make considerably less than that for a 15 person course even at Ivies. And they are doing the job of a grad student, not an adjunct. The lectures and materials are already there. They only have to grade things that can’t be graded automatically, like essays. All class participation can be recorded, making grading of that more fair and accurate.

This system is infinitely flexible. Want smaller sections? Pay more. More time? Pay more. 5 person section plus 1.5 hours/week of prof time? $2500. Don’t need one on one attention? Pay less. Lectures and materials only? $150. This way you can spend disproportionate money on the classes that really matter to you, or where you really need help, rather than paying a high price for all of them. And the section leaders can decide precisely what they want to take on.

And, at that salary, my section leaders need not be grad students. I can cherry pick from adjuncts who have cut their teeth at other colleges because I am outpaying everyone.

Will adjuncts really want to be reduced to the function of graduate students? Won’t they want to create their own lectures and make their own courses? Great! Did I mention that I pay on a royalty basis?

Meanwhile, 10% of the proceeds should be plenty to run the thing and make a profit, if I’m running for profit, once I have any scale at all. You don’t think so? 100,000 FTE students taking 8 $1000 courses a year? $80 million. How much do you think it costs to run a website to accommodate this for 100,000 students? $2 million would mean you are throwing a lot of parties. For all the dark warnings here about administrators here, it is they who should most fear this system. Why would a MOOC need any administrators at all? Unless accreditation mandates them, in which case accreditation standards should be revised.

At this point, all the work in complaints about in-person interaction is being done by “in-person”. You can’t just run into your professor in the quad, that’s true. The whole thing becomes more formal. But you also get an half-hour a week one-on-one, more if you like. That’s not nothing. There are real losses here, but there are also gains, and people seem to want to evaluate it looking only at the losses.

But of course some of these MOOCs are capitalist enterprises. Why would they pay so much to section leaders instead of keeping the money as profit? Well, because they have to compete with me in an open high-information market with relatively low cost of entry, and they will be selling an inferior product. Ironically, the one danger of this would come from established, especially highly-reputable players. Yale can put any junk in the world online, and, so long as they will provide regular Yale credit for it, they can charge a lot of money for it. They are Yale. Over time, this will degrade their reputation, but it could make them a lot of money in the meantime. I don’t think they will do this, but they could. But the only way Coursera is going to get their degrees respected is to earn it. This thread shows how skeptical an audience they are playing to.

This doesn’t address the socialization issue, or the entire complexities of the college experience. It describes a pure teaching institution, not one that engages in research. Those are more complex issues. But let’s at least nail down the basic question of what this economic structure makes possible. Eliminating the physical campus, eliminating virtually all need for administration and replacing it with a modest technical staff, and reducing lectures to media that can be presented at effectively no marginal cost – these things mean that what you are paying for is faculty attention, and you can therefore get much more of it for the same cost or lower cost.

No other savings than real estate? Really? Online courses are still going to require campus police forces? And security guards? They’re still going to require building maintenance and campus vehicles and gardening and painting all the eye candy? An onsite gym, a parking authority with a permitting system and parking people to enforce it? They still need onsite health clinics? They’re still going to require planning, building, etc permits for all construction and for any changes? They’re still going to maintain Student Unions and all the campus activities? They’re still going to need an office with an administrator and at least one administrative assistant for each department?

Royalties are calculated by computer. Wages paid by direct deposit. Most of what you’re talking about is automated in this day and age. I can write the code to handle my complex fee structure just fine, thank you. You can basically do it all in the database, other than user interface. You collect fees with a webpage that takes debit or credit cards. If you take financial aid, you have to interface with that system, which is probably somewhat complex, but once you build the interface, it should be pretty automated. If not, the system needs to be changed. After all, lenders interfacing to Fannie Mae, even for buyer qualification – automated; calculation and application of Medicaid benefits, once eligibility is determined – automated. If Pell Grants are behind, they need to catch up. Enrollment is pretty simple unless you want to be selective. There are advantages to that so it may be worth doing. Hiring and firing takes something, but you don’t seem to appreciate how much play there is in my numbers (and yes I did exaggerate the cheapness a bit for effect, but this idea that online will not save huge sums of money is ridiculous). Suppose I am off by a an order of magnitude in cost of operation. Barely makes a difference. Two orders of magnitude means I would have to charge about $300 more a class – still vastly cheaper than traditional alternatives. And that means I’m paying $200 million to run a website for 100,000 students and maybe 10,000 faculty.. But if I had to, I could afford that just fine. The high cost of university is physical plant, plus an established way of doing things, plus other considerations than teaching, primarily research, residence (though dorms themselves, I would imagine, are a net revenue source), entertainment, and the like. An office, an administrator, and at least one admin for each dept – that’s not exactly real estate, but it’s not something you need for a web enterprise. I’ve seen how Universities operate, though only from a student level, and I’ve also seen how web-based corporations large and small operate, from the inside, and the latter are much more efficient. We are talking here about the University as a web-based enterprise.

Instead of starting with a contemporary University, which presupposes a lot of ways of doing things that may not be optimal but merely established, start with Udacity or Coursera and ask what they need to add to function as I’ve described. They will have to add some things, to be sure. But there’s nothing they should have to add just because that’s how brick and mortar Universities do it, unless that is enforced by accreditation standards, in which case the standards should be changed.

Chris, never heard of OU. That’s even cheaper than what I had. Do you do a lot of small group and one on one interaction online?

Alex K, one of the joys of that article about the internet being humbug is that it was written for Newsweek!

Chris, you mean someone has to write the video? Sure, but just once – not every time you use it. The writing has no marginal cost. Do you have a physical office, an administrator, and an admin for each subject you offer a degree in at Open University?

Chris, it seems to me that the 10 million I mentioned is pretty good compensation for someone to write 300,000 words, put a voice over on stock footage, and indeed do considerably more than that. If I’m doing high production value in house, then I probably keep more for that and give less royalty, but it hardly matters: with 100,000 FTEs paying $1000 a course (average) and giving 80% to section heads, I’ve got $160 million dollars to play with. I can afford tons of office space, equipment, I can do research into education, but no way do I need to take on all the expenses of a physical university. That’s what I’m saying, not that I will not have an institutional structure at all. I think my office space – probably more bullpen style – would be divided more by function than academic field, but nothing hinges on this: I’m not going to suffer from being unable to afford office space, and, if doing it by subject works better, fine.

As for Silicon Valley being incapable of generating lasting institutions, you guys are not exactly the Catholic Church. Perhaps you feel you will outlast Google, but it is far from clear that you will.

Chris, so you have curriculum development as 30% with economies of scale, presentation and overhead at 70% with no economies of scale. I had presentation featuring mostly direct faculty interaction and other things faculty are employed to do (like grading) at 70%. But I’m also paying adjunct faculty 3 or 4 times what they usually get, so I have a lot of room to lower that if I need to put more into overhead. I’m also charging somewhat more than you, although much less than traditional private universities (and public ones of comparable quality have comparable costs, just less of it paid by the student). I didn’t see curriculum development as needing hundreds of millions of dollars, but I built into my estimates ridiculous amounts of wiggle room, dance room really.

But I don’t think you get what I’m doing here. The arguments that have dominated this thread are: 1) Online education is inherently much inferior and therefore can never take the place of traditional universities, and 2) Online education will not actually be cheaper, or much cheaper, than traditional universities, because the savings are actually quite limited. I take it you do not agree with either of these propositions. Neither do I. In the course of trying to get people to see how far off base they were, I came up with a hypothetical plan as an example. This is not a business I have founded. It is not even a business plan I worked on for a month. It’s something I thought up in about the time it took to type it. And so far the main criticism has been Collin’s: that I did not allow enough for conventional business overhead (Collin seems to have been lucky enough never to work at a company that heard of outsourcing. In my world, no one does their own payroll anymore. I suppose Universities do, though) And your that I did not allocate enough to specifically academic overhead and curriculum development (though I have scale working for me there, and persistence). But I had $78 million is revenue that I did not even allocate – could be profit, if I’m for profit, though I’m skeptical of for-profit universities (but that is a different issue, so I set it aside), but also could go to curriculum or overhead. Not enough? I don’t really have to pay adjuncts $12,000 a course. I think they are underpaid, and I would like to cherry-pick, but the market will actually bear much less than that and still let me cherry-pick. But I’m pulling stuff out of the air as the barest proof of concept, and you seem to object that the existing institution you work with has a more detailed estimation of how it works. I certainly hope so. You are in serious trouble otherwise.